This proposal is concerned with the development of mathematical models of the human network. Despite the collection of considerable experimental data on idiotypic interactions it has been difficult to fully appreciate the role of the network in the functioning of the immune system. Because the network is intrinsically an in vivo phenomenon possibly involving interactions among as many as 10(5) clones, in vitro experiments fail to provide answers to the most crucial questions. We propose to develop and analyze, using modern large scale computing techniques, models of immune networks with the ultimate goal of understanding the role of networks in immune regulation. We believe networks have not been understood because it is difficult to comprehend in the absence of mathematical models the operation of a large system of elements that interact according to nonlinear dynamical laws. As has been learned in other areas of science, it is virtually impossible to make predictions about large nonlinear complex systems on an intuitive basis. There have been a number of modeling efforts involving immune networks. We propose to utilize a new generation of models that aim to confront some of the complexities of the system. Among the issues we plan to explore are these: (i) The conflict between clonal selection and network views of the immune system. We believe that our models will show the immune system may be organized such that some clones are connected into a network whereas others remain isolated from the network. Thus clonal selection and networks both may be correct views. (ii) A perceived difficulty of network views is that it leads to the prediction that stimulation of the immune system by antigen can lead to perturbations involving the entire immune system. We propose to show that responses can remain localized within a network and thus effect only a limited number of clones. (iii) The nature of memory in the immune system. We propose to show that unlike neutral networks, dynamic memory may be stored in local regions of the network. (iv) The development of the network during ontogeny. Preliminary results indicate that as networks develop they self-regulate and develop a characteristic size and degree of connectivity. Competition for entry into the network can mold the immune repertoire of the adult animal. Specific health related issues include a proposed investigation of T cell vaccination for autoimmune disease. Most important is that progress in developing comprehension models will provide an increased understanding of the operation of the immune system as a whole in fighting disease. We think that it is due to our lack of understanding, and not necessarily a lack of knowledge, that we have generally failed to develop adequate therapies for autoimmune disease, allergies and other immune system disorders.